Unofficial JIFX

Introduction

The tenth meeting of the Research and Experimentation for Local and International First Responders (RELIEF) experiments took place from 29 Feb – 2 March at Camp Roberts in Paso Robles, CA. Over 70 people from 21 organizations in the disaster response community participated in 11 planned experiments.

Because the structure of RELIEF allowed for ad hoc experiments, participants created 5 unplanned experiments that tested the interoperability of their platforms.

Participating Organizations

• Burners without Borders. An NGO that builds sustainable communities in Peru and Haiti.
• CalState Long Beach GIS Department.
• Gatewing: the manufacturer of the small (sub 1m) UAV x100 and associated software platform.
• GeoIQ: a geospatial data management, visualization and analysis platform providing collaborative, browser-based data analysis tools for use by both technical and non-technical users.
• Harvard Humanitarian Initiative. The humanitarian interfaculty initiative at Harvard University.
• Lockheed Martin Corporation Universal Communications Platform: From Lockheed Martin, a system that integrates all types of fixed and mobile radio systems and enables interoperability with other communications and data-related systems.
• MedWeb: a telemedicine system used extensively by NGOs and the military.
• MutualLink: a company that builds peer-to-peer tools to connect disaster responders.
• National Geospatial-Intelligence Agency (NGA). Teams from R3 and the State NGA Support Team (NST).
• National Defense University: the co-host of the RELIEF event.
• Naval Postgraduate School: the host of the RELIEF event.
• NPS Remote Sensing Center. A new research center at NPS focused on the basic and applied research on remote sensing.
• Naval Surface Warfare Center – Dahlgren Division: the OSD- and Joint-Staff-sponsored QuickNets team, building a deployable crowdsourcing application for HADR/SSTR applications.
• OpenGeo. A company that builds open-source geospatial tools.
• Range Networks, Inc.: a California corporation that builds OpenBTS, an open-source software cellular system.
• Synergy Strike Force: a non-profit international network of synergists who facilitate information sharing in Afghanistan and other HADR/SSTR contexts.
• Team Rubicon. A veteran-run volunteer-based NGO that provides disaster relief.
• Tethr. A company that builds deployable crisis mapping tools.
• US Agency for International Development (USAID)/Office for Foreign Disaster Assistance (OFDA)
• US State Department Humanitarian Information Unit. The office at DoS under INR and the Office of the Geographer that analyzes humanitarian emergencies.
• Walking Papers/Field Papers. An RRTO-funded project to bridge the paper-digital divide for maps.

Tracks

Work during the experiment occurred within four planned tracks. Each track and experiment is explained in detail in the sections that follow.

1. Imagery Collection and Processing
• Aerial Field Spectroscopy of Tarps in Post-disaster Settings – NPS
• Structural Damage Assessment Utilizing Aerial LIDAR – NPS
• Disaster Imagery UAS – Gatewing
• Nextview Imagery Sharing – NGA
• Ad Hoc: Spectroscopic Change Detection with X100 Imagery – NPS-RSC & Gatewing NV
• Ad Hoc: Stalker UAS Video to Mutualink Nodes – Lockheed Martin & Mutualink

2. Mapping
• Field Papers – Walking Papers
• Ad Hoc: USAID/OFDA Integrated Web Mapping Technology – USAID, DoSHIU, NGA, OpenGeo

3. Communications and Crowdsourcing
• Lightweight and Resilient Cellular Networks – Range Networks, LLC
• Cellular Situational Awareness and Control – Lockheed Martin
• Mobile Operations Center – Lockheed Martin
• IRAP Interoperability – Mutualink
• Crisis Mapping & Communications – Tethr
• Ad Hoc: 2G Repeater for Sofcoast Mako – Lockheed Martin & Sofcoast
• Ad Hoc: 2G GSM Pico Cell for NetWarrior – Lockheed Martin & PM Soldier

4. Infrastructure as a System (various organizations)

Track 1: Imagery Collection and Processing

Reducing the time lag between the collection of fresh overhead imagery and making it useful to disaster responders has been a tough problem. Imagery cannot always be collected immediately. Some sources of imagery cannot be released quickly or in ways that permit wide distribution, not just because of classifications, but because of intellectual property licenses, download/transfer times, and the slow speed of the processing workflow.

The experiments around imagery at RELIEF 12-02 addressed several of these challenges.

Aerial Field Spectroscopy of Tarps in Post-disaster Settings (NPS Remote Sensing Center)

Making accurate counts of family units of displaced populations or refugee camps is a time consuming and difficult problem. However, distribution of tarps during humanitarian operations is an expected form of aid. Tarps are strongly correlated with family units as shelter. This experiment started building a spectral library of various tarps so that counting tarps might be automated by remote sensing professionals, speeding estimates of the number of displaced persons in a given area of interest (AOI).

Using tarps obtained from UN agencies and NGOs, the NPS Remote Sensing Center (RSC) team set out tarps at two locations: the Ranch House and at the airfield. They took spectroscopic pictures of the tarps to measure key attributes that they hope to feed into an automated detection algorithm. The team partnered with UAVs to collect overhead imagery of the tarps. The team also requested that NGA collect imagery over the site for later analysis to calibrate space-based imagery.

Structural Damage Assessment Utilizing Aerial LIDAR (NPS Remote Sensing Center).

After a disaster, organizations working on the unclassified side of the civ-mil divide often lack anything but the most coarse resolution elevation data (at best 10m and often only 30m SRTM), and almost no data that is useful for determining the damage to buildings. As was shown in Haiti, overhead imagery can be combined with Light Detection and Ranging (LIDAR) tools to greatly accelerate assessment of damage to structures. This experiment explored the application of LIDAR to structural assessments after a natural disaster.

Because the old barracks in cantonment area of the base area mix of standing and partially collapsed structures, this area lent itself to study by LIDAR. The NPS RSC obtained 2010 aerial LIDAR of the area and draped it over more current NAIP imagery. The RSC asked volunteers to identify damage from the three-dimensional image product. The results were promising, but demonstrated the need to have simultaneously (or close) collection of the imagery; some of the buildings had collapsed after the 2010 LIDAR collection, and this affected the results. We will look into repeating the experiment in August.

Disaster Imagery UAS (Gatewing).

The application of small UAVs to surveys of post-disaster areas is a subject of active investigation at TIDES. The Gatewing small X100 UAV explored how to integrate an existing survey UAV for application in disaster scenarios. The bird collected 3-16cm effective imagery (post-processing depending on altitude of collection) over the airfield as well as the area near the Ranch House. The flights supported several other experiments, including the NPS RSC experiment on tarps and an ad hoc experiment on change detection. The Gatewing team demonstrated a bird that could be preprogrammed to collect imagery over a small area (less than 2km2) with a team of 2 (pilot and observer).

Nextview Imagery Sharing (NGA, State HIU, USAID/OFDA).

For the past five RELIEF experiments, NGA and the US State Department have been exploring how to release commercial imagery purchased by the US Government to volunteer geographic organizations like OpenStreetMap. After reaching an agreement on the legality of the workflow in RELIEF 11-04, NGA and the State Department reconvened to work out standard operating procedures with OpenStreetMap. The outputs from the event included the Imagery Request Document that digital volunteers must use to request imagery from the USG; a plan for the imagery request service at the State HIU; and a set of requests for changes to the Humanitarian OpenStreetMap Team’s Tasking Server.

Ad hoc Experiment: Spectroscopic Change Detection with X100 Imagery (NPS-RSC & Gatewing NV).

Because the Gatewing UAV and the NPS Remote Sensing Center met at Camp Roberts, they explored means to use UAV-collected imagery to improve change detection algorithms. They imaged an area before and after setting up the tarps for experiment #1, thereby creating a data set with known and calibrated spectral properties for further analysis.

Ad Hoc Experiment: Stalker UAS Video to Mutualink Nodes (Lockheed Martin & Mutualink).

Mutual provides a peer-to-peer communications system (essentially a specialized VPN with a disaster response interface). Lockheed and MutualLink collaborated to channel a video feed from the Stalker UAS through the MutualLink peer-to-peer network to several sites in California and New Jersey.

Track 2: Mapping

While many people can rely on imagery and increasing resolution (pixels per given geographic area) to provide situational awareness, planners and logisticians need to turn imagery into maps and other cartographic products that show pathways for aid. The RELIEF experimentation teams confronted two key problems that have emerged since Haiti:

1. Bridging the paper-digital divide. Disaster responders prefer technologies that are low-power, durable, light, and easy to use. For all the power of tablets and smart phones, they still require power and a communications infrastructure. They are also costly to give as leave-behinds to the affected population. Paper meets all requirements, yet it has a significant disadvantage: transforming paper-based inputs takes valuable staff time. There needs to be a better means to bridge the paper-to-digital divide.

2. Connecting formal US Government systems on the tactical and strategic levels with volunteered geographic information. As Haiti demonstrated, volunteered geographic information (VGI) can become an invaluable asset. However, the policy and technical means to connect the US government with VGI is still nascent, especially on the tactical level of a forward operating base. There needs to be interfaces to the crowd that enable local commanders and DC-based policy makers to connect to fast-cycle VGI.

The experiments in this track addressed these problems separately.

Field Papers

On funding from the DoD/OSD/RRTO, Stamen Design and Todd Huffman developed the original Walking Papers concept into a more robust Field Papers. This project embeds geographic information onto paper maps through the use of 2D bar codes (QR codes) and an associated method for scanning any notes on the map via cell-phone cameras, scanners, or other available technology. The resulting scans can be fed back as base layers in common GIS systems.

Ad Hoc Experiment: USAID/OFDA Integrated Web Mapping Technology:

USAID/Geocenter has been tasked with building a web mapping capability. Because of proximity of several projects that provide web mapping tools, USAID/OFDA was able to explore integration of Oracle geospatial backend with several methods for meeting their emerging Geocenter requirements.

Track 3: Communications and Crowdsourcing

Communications is critical infrastructure during disasters. During this RELEF experimentation cycle, teams focused on restoration and interoperability of one of the tools most able to connect responders with affected populations: cellular communications.

Four planed and two ad hoc experiments integrated small (village-level) cellular systems with UAVs and dirigibles, and bridged these communications with other radios, including tactical handheld radios.

Lightweight and Resilient Cellular Networks – Range Networks, LLC

During last cycle’s experiment (12-01), Range Networks worked with MedWeb and setup their OpenBTS cellular system using a self-contained, solar-powered 1w cellular base station. This round (12-02) the team expanded on that work to explore integration of multiple small cellular systems under two scenartios: 1) when network backhaul must be made over satellite connections (BGAN and/or VSAT); and 2) when the cellular network might use UAVs to provide temporary coverage over a remote area.

Both configurations require a solution to two core routing problems: when the backhaul connection is using a private IP address and data gets routed via satellite, 1) how can multiple actors in the theatre constantly heal changes to IP allocations within the internal networks, and 2) how can the outside world reach individual devices on the network. The OpenBTS solution was (in essence) to create an analogue to the domain name server (DNS) system for all the devices on the cellular network. A tool tracks the internal and external IPs for all SIP endpoints on the system.

Using OpenBTS’s existing design of creating SIP endpoints between all handheld devices on its network, Range Networks attempted a wide integration of tools, including Ushahdi, Asterisk, Tethr, Tropo, and MedWeb over Wifi and 2.5G cellular using both VSAT and BGAN for backhaul. The important point is this: All the elements on the local side of the BGAN form a complete network of GSM/WiFi/VoIP/SMS.

In two experiments—one with Tethr supporting a local Ushahidi instance for crowdsourcing, and another for MedWeb integration to support the telemedicine platform used on 40 grey hulls—Range Networks developed the following architecture to connect simple message system (text/SMS), the Ushahidi crowdsourcing platform (via the Tethr experiment below), interactive voice response (IVR), and voice integration across a range of telephonic technologies.

IP Routing Diagram for Tethr Experiment
figure to be posted

VOIP Routing Diagram for Tethr Experiment
figure to be posted

IP Routing Diagram for MedWeb Experiment
figure to be posted

VoIP Routing Diagram for MedWeb Experiment
figure to be posted

The team learned several valuable lessons during both experiments:

1. Close proximity to radar systems used in the JIFX experiments caused VSAT issues which took more than day to resolve.
2. DTMF is not a reliable method of connecting Asterisk VoIP/IVR servers.
3. Multiple OpenBTS sites needs to be greater than 1km apart.

For more information, see https://wush.net/trac/rangepublic/wiki/RELIEF12-2#ActualEvent

Cellular Situational Awareness and Control – Lockheed Martin

Lockheed Martin explored how to use cellular networks to perform direction finding and filter/route all SMS messages inside a cellular network to provide greater situational awareness. The experiment had planned to have five parts: a) Demonstrate Direction Finding (DF) for any cellular device to locate both first responders and victims, b) Demonstrate the ability to capture all SMS text messages, c) Demonstrate the ability to detect cellular hardware configurations, d) Demonstrate the ability to disseminate specific messages to target groups, e) Demonstrate capabilities to provide seamless voice across heterogeneous communications networks to all participants in a humanitarian relief scenario.

At the experiments, Lockheed automatically captured all GSM cell phones in its bubble and was able to retrieve/intercept SMS messages on specific test phones. It also set up SMS groups (analogous to email list servs), and used the rerouting capabilities to connect responders and the affected population. Direction Finding capability had to be postponed due to a hardware availability issue.

Mobile Operations Center – Lockheed Martin

As with previous RELIEF experiments, Lockheed established a mobile operations center (MOC) at McMillan airfield with a 2G/3G cellular bubble. This cycle, Lockheed was able to send video feeds from both a Stalker UAS as well as smartphones working on 2G and 3G back to the MOC as well as to handhelds working in the area. NB: video is not usually possible over 2G data connections. The integrated communications package also bridged tactical radios, cell phones, VoIP, and the video from the UAS into one situational awareness package for local commanders.

IRAP Interoperability – Mutualink

Mutuaink provides a peer-to-peer infrastructure for disaster responders to share virtual private spaces for collaborating across IP, radio, video, and telephone. The team deployed an instance of the peer-to-peer network to connect the New Jersey Institute of Technology, the California IRAP community (including the San Francisco Police Department and the SF Pyramid Building). This initial demonstration of capability enabled the Mutualink team to connect a UAV video (via Lockeheed’s Stalker platform) feed into the peer-to-peer network for the first time.

Crisis Mapping & Communications – Tethr

Tethr provides a deployable, self-contained instance of Ushahidi, combined with a communications package that provides a local WiFi cloud connected to backhaul over either BGAN or 3G/4G cellular data systems. Tethr experimented with integration with the Range Networks OpenBTS system to explore how to received SMS calls from the OpenBTS network and integrate directly into a local instance of an Ushahidi crisis map. The experiment resulted in partial success: the Tethr team did integrate inbound messages from the OpenBTS system, but were not able to register those SMS reports directly into Ushahidi. This task will continue between RELIEF experiments.

Ad Hoc: 2G Repeater for Sofcoast Mako – Lockheed Martin & Sofcoast

SOFCoast provides a small, tactical dirigible (aerostat) called Mako, which is capable of small (2kg) payloads. The Lockheed Martin explored the possibility of using the aerostat to lift a 2G cellular repeater. The teams were able to get the power system working and to pass all pre-flight checks; however, weather prevented launch and the experiment was not able to fly during this RELIEF.

Ad Hoc: 2G GSM Pico Cell for NetWarrior – Lockheed Martin & PM Soldier Warrior

Lockheed provided 2G cellular coverage to the PM Soldier Warrior training occurring on the front side of Camp Roberts, including both voice and data connectivity.

Track 4: Infrastructure as a System

For the past eight experiment cycles, the TIDES project has focused its Camp Roberts implementations on information and communications technologies. During RELIEF 12-02, TIDES hosted a conversation to broaden its experimentation into the other infrastructures that are core to TIDES’s mission: power, water, sanitation, heating/lighting/cooling, integrated cooking, and shelter.

One of the core problems with physical “hard” infrastructure is that its design and deployment tends to be made by function. Water projects are viewed as separate from power. Sanitation separate from shelter. However, in a community, these infrastructures must function in an ecosystem of other physical processes. In addition, many hard infrastructure projects have failed because they did not integrate with the soft cultural infrastructure of a given place and time.

TIDES hosted a conversation of experts from civil affairs, special forces, logistics, and NGOs to raise questions about what TIDES should explore over the coming year at its experimentation cycles at Camp Roberts. Two insights of particular import emerged from the conversation:

1. Cross-functional problem solving teams are still needed in the field. Civil affairs and special forces used to provide teams who could mentor affected populations in a wide range of technical and planning skills. While USAID DART teams still have some of this character embedded in them, the DART teams are too few in number to make a significant dent in large SSTR activities. There is an open requirement for teams that can be sent to communities who can teach how to stitch various infrastructures into a functioning ecosystem that is tied to the cultural norms of the place and time.
2. We need to learn how small organizations are building infrastructures as systems so we can learn to scale the good ideas. There are projects which use systems thinking to design ecosystems of infrastructures. However, these projects are not well characterized in the military literature, and certainly not reflected in the TTPs for SSTR operations. There is a need to study the problem to understand what we are missing and how to move forward.

TIDES will take these insights into the May RELIEF 12-03 experiments, which will run 21-23 May 2012 in Arlington, VA.

Addendum: Team Rubicon Support

Team Rubicon is a volunteer force of veterans that respond to disasters and humanitarian emergencies. Two members of the team were observing the Camp Roberts experiments when tornadoes hit in the Midwest. Lockheed Martin provided the team with bandwidth and support while they mobilized their teams to respond.